11 Exceptional Speakers Contested and Problematized Identities

KIRA HALL

1. Introduction

The field of has witnessed several pivotal shifts in its interpre­ tation of normative and non-normative gender identity. This review aims to expose these shifts in an examination of the ways in which scholars have supported theor­ etical claims by referencing the speech patterns of "the linguistic deviant" - the speaker who fails to follow normative expectations of how men and women should speak. What becomes apparent in an overview of the literature is that linguistic deviance takes as many forms as the field has theories. In foundational discussions of language and gender in the early 1900s (e.g., Jespersen 1990 [1922]) the linguistic deviant is the "woman" herself, whose speaking patterns are peculiarly divergent from more normative (in this era of scholarship, male) ways of speaking. In early feminist work by those arguing for what has been termeddominance a model of lan­ guage and gender (e.g., Lakoff 1975), which theorizes women's speech patterns as a by-product of male dominance, the linguistic deviant is multiplied in some texts to include all speakers who are in some way disenfranchised from institutionalized male power - women, hippies, homosexuals, and even academic men. When the field shifted in the 1980s to adifference or two-cultures model of language and gender, which works on the assumption that children are socialized into divergent inter­ actional patterns within single-sex playgroups (e.g., Maltz and Borker 1982), the linguistic deviant resurfaced as tomboy and sissy, whose preference for other-sex playmates was discussed as proving the more normative, two-cultures rule. This latter use of the linguistic deviant could be said to parallel early discussions of non-Indo-European "women's languages" and "men's languages" in the first half of the twentieth century (e.g., Chamberlain 1912; Flannery 1946), where the "effem­ inate man" or "mannish woman" appeared in the footnotes as strange and deviant exceptions to an otherwise unshakable linguistic dichotomy.

The Handbook of Language, Gender, and Sexuality, Second Edition. Edited by Susan Ehrlich, Miriam Meyerhoff, and Janet Holmes. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Contested and Problematized Identities 221

Because the overwhelming majority of our field's theories have been based not just on the speech patterns of heterosexuals, but also on those of white middle-class English speakers, the deviant "ethnic" is also a common character, particularly in discussions that seek to make universal claims about how women and men speak. Most notable in this respect are studies supporting a two-cultures model of lan­ guage and gender, where women whose speech styles do not conform to those identified for the unmarked middle-class white woman become problematic for the theory. When scholars began to diversify the canon by studying the speech patterns of men and women in a variety of communities, societies, and cultures, a new theory of language and gender was born that had as its focus organizations of language and gender in communities of practice (Eckert and McConnell-Ginet 1992). It is only when the field shifts to this perspective that we begin to see the pur­ ported linguistic deviant on her or his own terms, as a member of a community whose speaking styles are influenced by more localized norms of language and gender. Because what is "normative" becomes potentially infinite within this theo­ retical paradigm, the research canon becomes diversified as well, enabling not only more sophisticated research on language, gender, and ethnicity, but also the devel­ opment of a field that has the sexual and gender deviance of previous generations at its center: queer linguistics. This chapter serves as what we might call an "underbelly" review of major works in language and gender research. It is not my intention to criticize earlier studies for their exclusions of certain communities of speakers, as all theories are limited by the intellect of the time in which they were developed. Rather, I offer this review as an exposition of the historical shifts governing our field's understanding of normativity on the one hand and deviance on the other. What I illustrate here is that the concept of non-normative gender identity while addressed in the gender and language literature in a peripheral manner until quite recently is neverthe­ less foundational to the major theoretical perspectives that have developed within the field.

2. Footnote Effeminates and Feminists

The field's first exceptional speakers surface in a flurry of anthropological discus­ sions on sex-based "languages" that appeared at the turn of the twentieth century Early anthropologists and ethnographers, in their explorations of non-European languages and cultures, developed the twin concepts of "women's language" and "men's language" as a means of explaining the morphological and phono­ logical differences they observed between the speech of women and of men. It is appropriate to begin our discussion here, not only because the work of these anthropologists ushered in a long trajectory of intertextual discussion regarding the social origins of gendered ways of speaking, but also because their represen­ tations of non-Indo-European languages initiated a dichotomous understanding of normative linguistic behavior that remains surprisingly influential in the field today. 222 Identities

What many of these texts have in common is what I identify here as "foot­ note deviance" - the casual and cursory mention of speakers who, simply put, do not play by the linguistic rules. Because so many of these scholars were, in pre-Whorfian mode, discussing divergent patterns of speaking for women and men in "primitive" societies as reflecting and reinforcing a social configuration of gender unknown to more "civilized " European cultures, the unyielding nature of the dichotomy between women's speech and men's speech was repetitively emphasized, so much so that scholars regularly spoke of these gender-influenced varieties as "separate languages" (e.g., Lasch 1907). What results is the kind of rep­ resentation aptly identified by Sara Trechter (1999) as linguistic exoticism, -where non-European languages, and the cultures carried through them, are portrayed as having rigidly defined gender roles, even to the point of restricting the way peo­ ple talk on the basis of sex. Early portraits of languages like English are hardly parallel, for even when divergent patterns of speaking for women and men are acknowledged, as in Otto Jespersen's (1990 [1922]) piece on "The Woman," they are discussed more as a matter of individual choice, if not taste. And so we arrive at the long-standing distinction in the literature between "sex-exclusive languages" and "sex-preferential languages," with the first designation giving the impression of rigidity and coercion and the second of fluidity and choice. Even as scholars are presenting the "women's languages" and "men's lan­ guages" of various non-European cultures as rigidly dichotomized and mutually exclusive, they also make mention of the speakers who buck the system. The most popular of these deviants is the effeminate man, the cross-talker whose nonconformity to a sex-exclusive language model makes him not just a linguistic anomaly, but a social weirdo, an outcast. The fact that he is labeled as "effeminate" or "womanly" by the rest of society for using women's language is then held up as evidence for the extreme and unforgiving nature of the model. An early example of this approach comes to us from Alexander Chamberlain's (1912) review of "Women's Languages" which appeared in American Anthropologist. Chamberlain quotes a Caraya speaker as declaring that one of the researchers "was a woman" because "he pronounced the Brazilianword jacuba (a kind of drink), not saúba, as a man would have done, butsãkúbã after the fashion of the women" (1912, 580). The anecdote underscores the separateness of the two varieties, since a male speaker who crosses the linguistic divide will not just be seen as womanly or effeminate; he will actuallybe a woman. The resulting portrait of women's and men's language use is rigidly dichotomous, so much so that a speaker's use of the "other" variety changes his sex altogether in public perception. A more modern example of this same approach is found 30 years later in Paul Furfey's (1944) review entitled "Men's and Women's Language," which includes the following footnote as a quick aside: "Particularly interesting was Dr. Herz- field's observation that a man using a woman's expression would be considered effeminate" (1944, 223 n.). As the author offers no further explanation in the foot­ notes as to why this observation is "particularly interesting," the import of the com­ ment is clear only when read alongside the larger argument developed within the text. Furfey repeatedly suggests that the sex-based linguistic differences evident in many non-European languages point to a "consciousness of men and women Contested and Problemaiized Identities 223 as different categories of human beings" - one that is, in his own words, "bound up with a masculine assertion of superiority" (1944, 222). The implication is that the same sort of hierarchical consciousness does not exist in European cultures, a point behind Furfey's stated goals for writing the article: "The present paper will discuss divergencies in the language usages of men and women, a phenomenon which is barely discernible in the familiar languages of Europe, but which is not at all uncommon among primitive peoples," for which it "serves as a tool of sex dominance" (1944, 218). By avoiding any in-depth discussion of gender in Euro­ pean languages, Furfey's review works to exoticize the oppressive nature of gender in non-European cultures. The most exotic proof of this oppression is the linguis­ tic effeminate, whose use of women's language emasculates him to a position of powerlessness. The significance of this emasculation potential is also articulated in Regina Flan­ nery's (1946) article on "Men's and Women's Speech in Gros Ventre," albeit for a rather different reason. We find a slight shift of tone in this article, as Flannery appears to move away from previous representations of sex-based speaking styles as distinct "languages" with her use of the term "speech differences." Yet Flan­ nery also emphasizes the mutually exclusive nature of these gendered styles by detailing the place of the "mannish" woman and "effeminate" man in the com­ munity, who serve as a source of "mortification" and "shame" for their families. We later learn that it is this very mortification and shame that is accelerating lan­ guage loss in the more general population. Because children are afraid that they will be laughed at by older generations for being "bisexual" if they use the wrong gendered forms - knowing, as they do, "the connotations in the minds of older generations" (Flannery 1946, 135) - they choose to avoid using Gros Ventre alto­ gether by speaking only English. Flannery's argument is a historical one, and our footnote effeminate wins the dubious distinction of promoting language shift. The kind of evolutionary logic reflected in Flannery's discussion of language shift is evident in the majority of these early descriptions of men's and women's languages, which regularly contrast the "archaic" and "primitive" nature of sex-exclusive language systems with the modernity carried by sex-preferential systems such as English. A case in point is Otto Jespersen's (1990 [1922]) early discussion entitled "The Woman," in which he outlines the many different kinds of sex differentiation evident in the world's languages. An important fact that has gone unnoticed about Jespersen's article - now infamous in language and gender studies for its representation of "the woman" as the linguistic other - is the evolutionary logic betrayed by its organization. Extreme phonetic differences existing in non-European languages give way to "very few traces of sex dialects in our Aryan languages" (206) followed by only "a few differences in pronun­ ciation between the two sexes" (209) in contemporary English. The vocabulary and word-choice differences evident for the sexes in English, in contrast to the phonetic differences evident for the sexes in non-European languages, hold a more advanced position on the evolutionary linguistic continuum. This representation hinges on Jespersen's sociological explanations for phonetic divergence, with primitive tribes and early civilized peoples sharing a sex-based division of labor that resulted in different phonological systems for men and women. Modern-day 224 Identities languages like English do not have distinctive grammars for the two sexes since the age-old division of labor has, in Jespersen's understanding, only "lingering effects" (219) in the twentieth century. This teleological logic is also betrayed by the kinds of exceptional speakers Jespersen chooses for three of his four "time periods" in language and gender relations. We move from the young Carib-speaking man who is not "allowed" to pronounce the war words of men's language until passing certain tests of bravery and patriotism, to the sixteenth-century French-speaking effeminate who imitates women in his reduction of the trilled r, to the modern-day English-speaking feminist who imitates the slang of men. The gendered rigidity evident in the non-European languages mentioned at the beginning of the article gives way to a certain fluency in the European languages discussed later, with the crucial turning point being sixteenth-century France. It is at this juncture, suggests Jespersen, that the sex-based division of labor, with its rigid linguistic reflexes, is replaced by a sex-based public /private dichotomy - a sociological shift that leads not to separate languages, but to the possibility of cross-sex appropriation. Jespersen's exceptional speakers thus enter the text in order to illuminate how our present-day linguistic and cultural situation differs from that of the less civilized world that precedes us. The height of this linguistic evolution is captured by the educated feminist of the final time period. Her use of the "new and fresh expressions" of men, precipitated by "the rise of the feminist movement" (1990 [1922], 212), points to an equality between the sexes that was heretofore nonexistent. The divergent uses of vocabulary and syntax that Jespersen subsequently identifies are then theorized not as sociological, but as cognitive, psychological, and personal.

3. The Woman

Given the care with which many of these early anthropologists describe both "men's language" and "women's language" as normative aspects of a particular linguistic and cultural system, Jespersen's more concentrated focus on "the woman" marks an important theoretical shift in the literature. Jespersen ushered in a new understanding of linguistic deviance, with English-speaking women and their speech peculiarities usurping the cross-talking effeminates of non-European cultures. For Jespersen, men's speech is the norm and women's speech the deviation. This is a new form of linguistic exoticism, one that has "women's speech" in modern-day English as its target instead of the women's and men's languages of non-European cultures. The scholars who followed Jespersen, also observing differences between women's conversational patterns and the more socially accepted or dominant patterns of men, tended to represent women's speech as abnormal, as the marked case, as norm-breaking. In this segment of our field's early history, then, the most contested and problematized gender identity becomes "the woman" herself. Literature that addressed the historical development of women's and men's lan­ guages contributed to this perspective (see Haas 1964 [1944] for a review), as it Contested and Problematized Identities 225 tended to position women's forms as derivational, even in cases where they were thought to be more archaic. For example, Gatschet (1884,79-81) discusses the exis­ tence of an "ancient female dialect" in Hitchiti, still spoken by women and elders in the community. But even though he claims that this dialect was formerly the language of men as well as women, he chooses to provide a grammar only of the newer "common form (or male language)." Although the women's variety is older and apparently basic, Gatschet's prose positions it as both "uncommon" and marked. Similarly, Edward Sapir (1949 [1929]) is careful to argue for two different directions of derivation in Yana, with male forms fundamental in some cases and female forms fundamental in others. Yet in his conclusion, when theorizing why these sex forms might have come to exist in the first place, he ignores the latter of these directions and discusses women's forms as purely reductive (a decision that seems to rest on an earlier observation that the male form in both cases "is longer than the female form"): "Possibly the reduced female forms constitute a conven­ tionalized symbolism of the less considered or ceremonious status of women in the community Men, in dealing with men, speak fully and deliberately; where women are concerned, one prefers a clipped style of utterance!" (1949 [1929], 212). Women cannot win in these early texts: when their language forms are discussed as funda­ mental or older, they are theorized as conservative and archaic in comparison to their more innovative and youthful male counterparts; when their language forms are discussed as derived or newer, they are theorized as psychologically deviant or otherwise abnormal. The simple fact that so many of the early articles on sex differ­ entiation in language carry the title "Women's Speech" or "Women's Language" points to an understanding of male speech asthe language and women's speech as a kind of oddity The term "peculiar" becomes the most common descriptor for women's speech in the literature of this period. Jespersen (1990 [1922]) himself is a big fan of this buzzword, using it to describe women's divergent uses of vocabulary (e.g., citing Greenough and Kittredge 1901, 210, "The use ofcommon in the sense of 'vulgar' is distinctly a feminine peculiarity"), as well as to theorize women's divergent uses of syntax ("These sentences are the linguistic symptoms of a peculiarity of feminine psychology": 216). His prose parallels that of Bogoras (1922,665-666) in his article on Chukchee published during the same year, who also discusses certain facets of women's pronunciation as sounding "quite peculiar" and "not easily understood by an inexperienced ear" (665). Bogoras's discussion is an especially clear case of the male linguistic gaze that characterizes much of this literature, with the author assuming a male readership that would identify with male uses of the language as opposed to female ones (certainly these phonetic forms do not sound so peculiar to the women who use them).

4. Hippies, Historians, and Homos

We find reflexes of this early trend even in the ethnographically informed dis­ cussions of women's and men's speech patterns that surfaced with the rise of 226 Identities speech act theory in the 1960s and 1970s. Elinor Keenan (Ochs)'s (1996 [1974]) influential study of Malagasy speakers in Madagascar, entitled "Norm-Makers, Norm-Breakers: Uses of Speech by Men and Women in a Malagasy Community/' is a case in point. Keenan (Ochs) spends the first three-quarters of her article outlining the linguistic repertoire of "the people of Namoizamanga" (1996 [1974], 100), describing in great detail their varied discursive strategies for avoiding direct affront. It is only in the last few pages of the article that we come to realize that women are not included in this description because of their preference for a more direct and confrontational speaking style. Although Keenan (Ochs) presents Malagasy-speaking men as "norm-makers" and Malagasy-speaking women as "norm-breakers," the women of her study are certainly adhering to "a norm" just as much as the men are: their expected participation in more direct forms of information-finding, bargaining, and child-scolding speaks to the strength and persistence of that very norm. But since it is a norm deemed inferior by the more dominant male-speaking population, Keenan (Ochs) chooses to portray the speech of these Madagascar women as deviant, or even (as the title of her article might imply) subversive. The representation of women as a problematized gender identity, then, becomes central to feminists working within thedominance model of language and gender, which focuses on how women's speech patterns are trivialized, or otherwise marginalized, in male-dominant societies. Norms in such studies are viewed as singular, and women become the non-normative exception. But the women of these texts rarely stand as the lone exception to an oppressive discursive regime. As with Sapir's (1949 [1915]) work on deviant speech in Nootka, early researchers frequently discussed the speech patterns of women with reference to other marginalized identities in order to emphasize their abnormality, or as in the case of Robin Lakoff (1975), to highlight their disenfranchisement from the powers that be. Most scholars have read Lakoff's work as exclusively concerned with women's patterns of speaking, ignoring her rather extensive discussions of a variety of other identities, among them the effeminate homo­ sexual, the anticapitalist hippie, and the asocial male professor. For Lakoff, such men have much in common with women: specifically, they share a marginality determined by their exclusion from institutionalized male power. Central to Lakoff's explanation for this shared marginality is the gendered division of labor, and, more specifically, the differential orientations to politeness brought about by this division. Women, excluded from a male workplace built on "present-day reflexes of male bonding" (1975, 77), tend to orient themselves to politeness forms that discourage bonding, gravitating toward the first two rules of Lakoff's politeness paradigm: formality (keep aloof) and deference (give options). Men, on the other hand, as a result of their socialization within workplace situations that require them to develop techniques of working together as a group, are more likely to embrace Lakoff's third rule of politeness:camaraderie. The latter rule would be essential in, for example, a male-dominated corporate workplace, as group members must develop interactive measures to gloss over emotional reactions and disagreements that might hinder progress toward a common goal. These are measures women have generally not needed to develop, Lakoff suggests, since Contested and Problematized Identities 227 they have historically been excluded from these group-oriented work environ­ ments. Her remark that women's use of terms likedivine is "not a mark of feelings of inferiority but rather a mere badge of class" (1975, 52) is telling in this respect. Consistent with radical feminist discussions of the time that identified women as a fourth world (Burris 1973) orseparate caste (Dunbar 1970), Lakoff situates women within a powerless "female class" that exists outside of the institutionalized power structure and employs a non-work-related vocabulary deemed irrelevant by this very power structure. To borrow from the title of Lakoff s book, "woman's place" is a place excluded from the public sphere of men's work, and the lan­ guage patterns that have developed as a result of this exclusion are devalued as "women's language." The notion of a masculine workplace, then, is fundamental to Lakoff's theo­ retical explanation for men's and women's differential use of linguistic phenom­ ena. This explains why academic males, hippies, and homosexuals occupy the margins of Lakoff's text as problematized gender identities. Like women, these groups are in some way excluded from a social history of male bonding in the labor force, and as with women, this exclusion leads to language patterns disso­ ciated from what Lakoff terms "real-world power." The following excerpts from Lakoff's text underscore the fact that her work is not so much about gender as it is about power:

Hippies

I think it is significant that this word ["groovy"] was introduced by the hippies, and, when used seriously rather than sarcastically, used principally by people who have accepted the hippies' values. Principal among these is the denial of the Protestant work ethic: to a hippie, something can be worth thinking about even if it isn't influen­ tial in the power structure, or moneymaking. Hippies are separated from the activities of the real world just as women are - though in the former case it is due to a decision on their parts, while this is not uncontroversially true in the case of women. (Lakoff 1975,13) Academic men

Another group that has, ostensibly at least, taken itself out of the search for power and money is that of academic men. They are frequently viewed by other groups as analogous in some ways to women ... what they do doesn't really count in the real world ... The suburban home finds its counterpart in the ivory tower: one is suppos­ edly shielded from harsh realities in both. Therefore it is not too surprising that many academic men ... often use "women's language." (Lakoff 1975, 14) Homosexuals It is of interest, by the way, to note that men's language is increasingly being used by women, but women's language is not being adopted by men, apart from those who reject the American masculine image [for example, homosexuals]. This is analogous to the fact that men's jobs are being sought by women, but few men are rushing to become housewives or secretaries. The language of the favored group, the group that holds the power, along with its nonlinguistic behavior, is generally adopted by the other group, not vice versa. (Lakoff 1975,10) 228 Identities

For Lakoff, male hippies, academics, and homosexuals are all in some sense gender deviants - identities who have forsaken a capitalistic power structure built on mas­ culine ideals for pursuits considered trivial in the "real world." This would explain, suggests Lakoff, why the language patterns of hippie, academic, or homosexual so often appear to resemble that of the American middle-class housewife. That these disenfranchised groups are likely to use some of the same specialized lexical items as American middle-class women, she argues, points to a more general conclu­ sion: "These words aren't, basically, 'feminine'; rather they signal 'uninvolved' or 'out of power'" (Lakoff 1975, 14). While certain patterns of speech may be con­ sidered feminine because women are, in her own terms, the "'uninvolved' 'out of power' group par excellence/' Lakoff is careful to note that any group in society may use patterns associated with "women's language" (an observation that best explains her consistent use of scare quotes around the term). For Lakoff, it is the feminine-sounding male, marginal to the world of institutionalized masculinity, who ultimately enables her to formulate the crux of her argument: "The decisive factor is less purely gender than power in the real world" (1975, 57). In spite of their centrality to Lakoff's theory, these marginal figures are fre­ quently overlooked in subsequent discussions of her work. The majority of her critics, swept up in an imperative to test her argument empirically, interpreted Lakoff as suggesting that only women speak "women's language," developing study upon study to determine whether or not female speakers actually use this register more than their male interlocutors. What is amusing, in retrospect, is that a great number of these studies analyze the speech patterns of the very academics that Lakoff identifies as linguistically divergent in order to "disprove" her hypothesis, such as Dubois and Crouch's (1975) study of men's use of tag questions at an academic conference. But the mission of critics to distinguish the female register from the female speaker (see also Crosby and Nyquist 1977; O'Barr and Atkins 1980) is nevertheless admirable. It is this distinction that in many ways enabled the development of queer linguistics - a field that.explicitly questions the assumption that gendered ways of talking are indexically derived from the sex of the speaker.

5. Sissies and Tomboys

The 1980s ushered in an alternative flavor of language and gender research, marked in part by Daniel Maltz and Ruth Borker's (1982) proposal of a new framework for examining differences in the language use of American women and men. Their approach, sometimes identified as a two-cultures or difference model of language and gender, holds that American women and men come from two different sociolinguistic subcultures, in which they learn different rules for interacting with one another and interpreting conversational contributions. In an extension of John Gumperz's (1982) cultural explanation for interethnic (mis)communication, Maltz and Borker argue that boys and girls orient to their own sex as preschoolers and thus develop divergent cultures of talk. (See Contested and Problematized Identities 229

Goodwin and Kyratzis, Chapter 26 in this volume, for discussion.) The singular norm of studies in the dominance approach becomes dual again, with male and female speakers traveling on different (and frequently oppositional) tracks of normativity. Interesting in Maltz and Borker's platform is a short aside in their concluding notes, where they give us the "tomboy," together with "lesbians and gay men," as one of "a number of specific problems that appear to be highly promising for future research" (1982, 94). Why these marginal identities might be "potential research problems" for a two-cultures approach to language and gender is fairly clear. Because the argument is based on the assumption that boys and girls are socialized into interaction differently in their single-sex playgroups, what happens to the theory when we find children who appear to shun this very socialization? Do they grow up to be lesbians and gay men who share conversa­ tional patterns with the other sex? The sissy and the tomboy, then, as apparent exceptions to a socialization rule presented as having few if any defectors, become oddly important to a two-cultures perspective. The most overtly theorized discussion of sissies and tomboys appears in Eleanor Maccoby's (1998)The Two Sexes: Growing Up Apart, Coming Together, a comprehen­ sive review of previous research that supports a two-cultures approach to the sub­ ject of gender. Maccoby is interested in how biological, social, and cognitive forces come together to constitute what she calls gender's "explanatory web," creating divergent patterns of behavior for the two sexes that begin in the womb, material­ ize in early childhood, remain through adulthood, and are ultimately transferred to the next generation. In contrast to much of the two-cultures research that has as a main goal a description of "what boys do" as opposed to "what girls do" (offering linguistic evidence, for instance, to support the claim that boys7 interaction is more "hierarchical" while girls7 is more "collaborative77), Maccoby seeks to determine why these interactional differences arise in the first place. As her focus is on gender conformity in same-sex childhood playgroups, not dissension, tomboys and sissies appear in the text not so much as trouble-shooters for a two-cultures approach (or as identities whose interaction is interesting in their own right), but as exceptions that prove the more normative rule. And because this normative rule is produced biologically as well as socially for Maccoby, our tomboy and sissy come to play an interesting role in her theorizing of each of these influences. Maccoby's primary sociological argument for why divergent patterns of inter­ action exist between the two sexes has to do with the "greater strength77 (1998, 41) of boys7 playgroups as opposed to girls7. The forces binding groups of boys together, she argues, are much stronger than those binding girls together, leading to a much more exclusionary kind of play in which peer group acceptance becomes the overriding concern. Boys therefore have a much greater need for recognition from other boys, and this drives them to engage in the status-oriented discursive behaviors identified by many linguists for all-boys7 groups. The fact that girls do not enact sanctions against tomboy behavior in the same way that boys enact sanc­ tions against sissy behavior illustrates that boys7 groups are more cohesive, more conforming, more gender-exclusionary: "Clearly, an essential element in becoming masculine is becoming not-feminine, while girls can be feminine without having to 230 Identities prove that they are not masculine" (Maccoby 1998,52). It is worth noting that Mac- coby's use of the tomboy is diametrically opposed to Lakoff's (1975), who points to the "little girl [who] talks rough like a boy" as evidence for the strength of female socialization. For Lakoff, the fact that the tomboy is "ostracized, scolded, or made fun of" by parents and friends is suggestive of how society "keeps her in line, in her place" (1975, 5). The disparity between Lakoff's and Maccoby's sociological analysis of the tomboy could be a result of the 20-year time differential between the two texts. But to say that boys' groups are more cohesive because thesissy label operates as an insult whereas tomboy does not, as Maccoby does, ignores the import of age on peer acceptance of gender deviance. Certainly, Penelope Eckert's (2002; Chapter 27 in this volume) research on adolescent girls' management of the "heterosexual marketplace" suggests that it would be quite difficult, if not socially detrimental, for a girl to continue her tomboy leanings into the teen years. The differences of perspective voiced here undoubtedly have much to do with the fact that there is very little ethnographic, much less linguistic, research on so-called "deviant" gen­ der identities in either childhood or adolescence. The tomboy's unwritten nature, then, makes her ripe for all sorts of scholarly pickings. In fact, Barrie Thorne (1993), in her ethnographic study of gender in American elementary schools, discusses tomboys and sissies as part of a larger critique of the very two-cultures approach espoused by scholars like Maccoby, arguing that the variation we find within is greater than the variation we find between boys and girls taken as groups. For Thorne, the tomboy is just one aspect of a "complicated continuum of crossing" (1993,112) - a continuum that is, in her opinion, obscured by research that operates on the assumption of gender as separation and difference. Maccoby argues that gendered behavior in childhood is a function of biology as well as of socialization, so it is not surprising that we find extended discussions of prenatal deviants as well. We learn about the male play patterns of girls who were exposed to excess amounts of adrenal androgen while in the womb (identi­ fied in the scientific literature as females with adrenogenital syndrome, or AGS), along with the rough-and-tumble play of female rhesus monkeys whose mothers had been injected with testosterone when pregnant. Maccoby is careful to avoid drawing links between this scientific research and sociological discussions of actual tomboys, but here again we see deviance embraced as evidence for normativity. The argument goes something like this. "Normal" boys and girls, as a result of prenatal hormonal priming, have different rates of maturation when it comes to particular kinds of behavior. Girls appear to self-regulate their behavior much ear­ lier than boys do, having earlier success at potty-training, for example, and show­ ing faster progress in language development. A boy's lack of self-control earns him more hierarchical, disciplinary commands from his parents as well as more rough-and-tumble play; a girl's more advanced language capacity invites more relational and nurturant talk about feelings. These same children eventually come to self-select playmates who behave as they do. The resulting single-sex playgroups begin to accentuate the behaviors encouraged earlier by parents, until definitively divergent patterns of interaction emerge for the two groups. The AGS girl stands on Contested and Problematized Identities 231 the sidelines of this discussion, stepping in at critical junctures as evidence for the biological component of Maccoby7s explanatory web. The fact that AGS girls prefer male play partners and high levels of rough-and-tumble play gives Maccoby the evidence she needs to argue for biology's role in the construction of dichotomous gendered behaviors. And it is the biological aspect of Maccoby's argument that is particularly powerful, as it enables her to make a universal claim about how gen­ der operates. Our bio-tom, then, in her conjoined biological and social deviance, provides evidence not only for a two-cultures gender normativity, but also for its cross-cultural persistence. One last remark is called for regarding the way in which Maccoby suggests that the phenomenon of early same-sex attraction might have an additional evolution­ ary purpose. She references anthropologist Arthur Wolfs (1995) study of boys and girls in southern China who, because they had been affianced by their parents at an early age, lived together in the same household for several years in preparation for marriage. Wolf found that such children come to lack sexual interest in each other when they reach adolescence, offering as evidence the fact that their sub­ sequent marriages have exceptionally low rates of fertility. Maccoby7s interest in Wolfs research again has to do with the biological aspect of the explanatory web, as his findings provide yet another biologically oriented reason for why same-sex segregation might occur: "Children's spontaneous avoidance of cross-sex others who are not kin serves the biological function of keeping these others within the pool of potential mates77 (Maccoby 1998, 94). Yet this claim forces a connection between gender identity and sexual orientation. If tomboys and sissies spend much of their childhood with "the other77 sex instead of their own, do they then, as Wolfs theory implies, grow up to lack sexual interest in the opposite sex? Is this where lesbians and gay men come from? Certainly, Maltz and Borker's (1982) juxtaposi­ tion of "tomboys77 and "lesbians and gay men77 as potential problems in their early research platform implies some connection between early deviant gender identities and the sexual orientation of adults. Indeed, the conflation of gender and sexual identity appears through much of the language and gender literature, where, until quite recently, the conversational practices of lesbians and gay men are discussed not as indexing community membership, but as instancing gender deviance. A teii ing example of this conflation surfaces in Burrell and Fitzpatrick (1989), where we find the heterosexualization of a conversational excerpt that takes place between two gay men in Deborah Tannen7s That's Not What I Meant! How Conversational Style Makes or Breaks Relationships (1986). In her bestseller, which includes a chapter on the cross-cultural nature of male-female communication, Tannen gives us one of the field's first gay couples in the form of Mike and Ken, whom she describes, refreshingly, as "two people who lived together and loved each other77 (1986, 126). The excerpt at issue regards a fight over salad dressing, where, according to Tannen, each partner misunderstands the conversational frame used by the other. But while Tannen discusses this exchange in gender-free terms in order to demonstrate the kinds of misunderstandings that can occur in close relationships, Burrell and Fitzpatrick reinterpret virtually the same exchange entirely along gendered lines; Mike and Ken even surface as "Bob77 and 232 Identities

"Joanne." The two excerpts - Tannen's followed by Burrell and Fitzpatrick's - are reproduced below:

Mike: What kind of salad dressing should I make? Ken: Oil and vinegar, what else? Mike: What do you mean, "what else?" Ken : Well, 1 always make oil and vinegar, but if you want, we could try something else. Mike: Does that mean you don't like it when I make other dressings? Ken: No, I like it. Go ahead. Make something else. Mike: Not if you want oil and vinegar. Ken: I don't. Make a yogurt dressing. (Mike makes a yogurt dressing, tastes it, and makes a face.) Ken: Isn't it good? Mike: I don't know how to make a yogurt dressing. Ken: Well, if you don't like it, throw it out. Mike: Never mind. Ken: What never mind? It's just a little yogurt. Mike: You're making a big deal about nothing. Ken: You are! (Tannen 1986,119)

Bob: What kind of salad dressing should I make? Joanne: Vinaigrette, what else? Bob: What do you mean, "what else?" Joanne: Well, I always make vinaigrette, but if you want make something else. Bob: Does that mean, you don't like it when I make other dressings? Joanne: No, I like it. Go ahead. Make something else. Bob: Not if you want vinaigrette. Joanne: I don't. Make a yogurt dressing. (Bob makes a yogurt dressing, tastes it, and makes a face.) Joanne: Isn't it good? Bob: I don't know how to make a yogurt dressing. Joanne: Well, if you don't like it, throw it out. Bob: Never mind. Joanne: What never mind? It's just a little yogurt. Bob: You're making a big deal about nothing. Joanne: You are! (Burrell and Fitzpatrick 1989,176-177)

How Tannen's gay men wound up as heterosexuals in Burrell and Fitzpatrick's book is not entirely clear, but their transformation offers an illuminating example of how sexual identity is often disregarded within a two-cultures model of lan­ guage and gender. What interests me is how the authors reformulate the excerpt as a conversation between "the independent spouse" Bob and the "traditional wife" Joanne. "Throughout this admittedly trivial interaction," the authors explain, "the independent spouse, Bob, saw his wife as becoming increasingly more demand­ ing, whereas the traditional wife Joanne, perceived her husband as becoming more Contested and Problematized Identities 233 hypersensitive and temperamental'" (1989,177). That an excerpt between two gay men is so easily recast into a heterosexual discussion of "The Psychological Reality of Marital Conflict" betrays a much larger theoretical problem in the language and gender literature of the 1970s and 1980s: namely, the persistent assumption that sexual identity is really about gender. When gays and lesbians do receive mention in the two-cultures model, they tend not to be subjects of study in their own right, but tangential characters who provide extreme evidence for a dichotomous view of gendered behavior. Tannen, for instance, in her subsequent bestsellerYou fust Don't Understand (1990), refers to Philip Blumstein and Pepper Schwartz's (1984) popular finding that "lesbians have sex less often than gay men and heterosexual couples" as support for her argu­ ment that men tend to be initiators and women respondents: "But among lesbians, they found, often neither feels comfortable taking the role of initiator, because nei­ ther wants to be perceived as making demands" (Tannen 1990, 147-148). Here, lesbians come to serve as a test-case for Tannen's theory, providing an archetypal female-female example of the behaviors she identifies for women speakers more generally. Tannen reads the purported lesbian hesitancy to initiate sex as a gen­ dered trait, and offers it as evidence for a more general theory regarding women's discomfort with self-assertion. Lesbians, as same-sex partners, are discussed as a kind of "grown-up" version of the childhood all-girl playgroups so instrumental to two-cultures theorizing. Tannen's occasional comparisons of lesbians with gay men, as in a later chapter of the book when she contrasts lesbian and gay under­ standings of the relationship between money and independence (1990, 292), are intended not as discussions of sexual identity, but as paradigmatic examples of difference between women and men.

6. Queers and the Rest of Us

What is exceptional about Tannen's lesbians and gay men, however, is precisely that they are not exceptional; that is, their interactive behaviors are viewed not as deviant, but as entirely in line with the interactive behaviors of heterosexual women and men. While some scholars may fault her work for failing to consider the potential influences of sexual identity on conversational exchange (see Jacobs 1996), her refusal to portray lesbians and gay men as peculiarly deviant, in the manner of former generations of researchers, is better understood as progressive for the linguistic scholarship of the time. Her work might even be said to reflect a transitional point in the academic treatment of sexual identity, when identities previously viewed as deviant or non-normative began to be brought into the mainstream of scholarly discussion. I want to argue here that three theoretical moves in the language and gender research of the early to mid-1990s precipitated this transition: first, the introduction of the notioncommunities of of practice (Eckert and McConnell-Ginet 1992); second, the more sophisticated development of ideological approaches to the study of language and gender (e.g., Bucholtz and Hall 1995; Bucholtz, Liang, and Sutton 1999; Gal 1991); and finally, the birth 234 Identifies of queer linguistics (Livia and Hall 1997), a field that activates, albeit critically, the philosophical notion of performativity. All of these moves were formulated within, and influenced by, larger theoretical moves in the academy. Most notable in this respect is multicultural feminism, which encouraged the intellectual embracement of heretofore understudied identities in a postmodern drive to diversify the academic canon. The linguistic reflexes of this drive, accordingly, share a focus on more localized organizations of language, gender, and sexuality The two-norm approach of the previous generation gave way to a paradigm that reframes the normative as ideologically produced within specific practice-based communities. Norms of feminine and masculine speech, although always con­ strained and influenced by dominant ideologies of language and gender, become potentially infinite in local articulation, particularly as gendered ideologies are produced only in interaction with localized understandings of race, class, sexuality, and age. The concept of gender performativity, as developed within queer linguistics and more generally in sociolinguistics, is closely allied with ideological and practice-based approaches to the study of language and gender, although this fact has been little discussed in the literature. As Anna Livia and I argue in our introduction toQueerly Phrased (Livia and Hall 1997), the concept is much needed in the field as a way out of the circular research paradigm encouraged by the theoretical tenets of social constructionism. The feminist distinction between sex and gender, with the first term being used for the biological and the second for the social, was a politically necessary one, as it threw a decisive wrench in essentialist arguments that limited social agency to biological predisposition. But this distinction also had a compromising effect on ethnographic research, leading language and gender scholars, for example, to seek out the sociolinguistic reflexes of a prediscursive biological sex. Working from the assumption that the social maps onto the biological (a perspective criticized by Nicholson 1994 as a "coat-rack model" of sex and gender; see McElhinny 2002; Chapter 2 in this volume), researchers pre-identified their subjects as "male" and "female" and then isolated the conversational strategies that distinguished these groupings from one another. Sexual identity, as a subjective designation not easily related to biology, remains invisible within this paradigm. But the performativity of gender, as formulated by Judith Butler (1990) via a Derridean reworking of J. L. Austin's (1962) notion of the "performative utter­ ance," disallows sociolinguistic approaches to identity that view the way we talk as directly indexing a prediscursive self. To a poststructuralist like Butler, there is no prediscursive identity, as even our understanding of biological sex is pro­ duced through cultural understandings of social gender. This kind of thinking puts much more weight on the speech event itself, requiring us to examine how speakers manage ideologies of feminine and masculine speech in the ongoing production of gendered selves. It also gives us a nonessentialist understanding of personhood, as what becomes important is not how speakers affirm or resist a pregiven biological designation, but how they activate various identity positions within particular con­ versations and localized contexts. Rusty Barrett's (1999) work on the "polyphonous Contested and Problematized Identities 235 identity" displays of African American drag queens in a Texas gay bar is an exem­ plary model of how such research might proceed, as he illustrates the ways in which speakers make use of linguistic variables with indexical associations to a variety of social categories. Yet Butler's theory also has its limits for ethnographic sociolinguistic research. Most pressing in this regard is the restricted agency awarded the subject in a poststructuralist focus on discursive determinism, together with the undertheo­ rization of the local in a philosophical text concerned with universal explanations for how gender works. Here is where the field would do well to remember how Austin's performative was taken up by linguistic anthropologists such as Dell Hymes, Charles Briggs, and Richard Bauman in the early ethnography of speaking. While Butler focuses almost exclusively on the rigid regulatory frames that make femininity and masculinity intelligible (in Austinian terms, the "conventional procedures" that make a performative utterance felicitous), these authors focus also on the emergent properties of specific speech events (see Hall 1999). Hymes's (1975) repeated call to "understand structure as emergeñt in action" is critical here, as he and other scholars of performance, most notably Bauman and Briggs (1990), led us away from the analysis of ritual as mere reiteration. What moves into focus with their work is not Derridean iterability but "the total speech act," as they uncover not just the cultural conventions that make performance, ritual, and even everyday conversation felicitous, but also the creative aspects that govern any speech event. Butler's limitation of creativity to resignification - as, for instance, when a drag queen performs the "wrong" gender and thereby exposes the constructed nature of gender perceived as natural - is impoverished in ethnographic terms, since it reduces drag queen performance to an appropriation of a dominant ideology of femininity. This is, indeed, the assumption behind Butler's argument that drag is a kind of "double mimesis," that is, men acting like women acting like women. But as Barrett so cogently demonstrates in his linguistic research, drag queens are not acting likewomen; they are acting like drag queens. Their interwoven appropriations of African American Vernacular English, the "Standard" English phonology associated with "white woman style," and lexical items indexical of gay male speech suggest that gender identity is a multivocal phenomenon that depends on interaction with other social identities for its articulation. Because drag queen identity is always localized and produced through a variety of conflicting cultural scripts (race, class, sexuality, and gender among them), it would be ethnographically reductive to discuss their performances purely as a subversion of a nonlocalizable "femininity." This brings me to the crux of an argument about how Butler's theory of gender performativity must be reworked, or at least acquire new focus, in the sociolinguistic study of language, gender, and sexuality The only way identities previously regarded as non-normative can be brought into the mainstream of scholarship is if we localize what constitutes "felicitous" and "infelicitous" performances of gender and sexual identity within the language ideologies circu­ lating in specific communities of practice. To discuss drag queen performance as 236 Identities the infelicitous enactment of dominant conventions of gender, as Butler does in her focus on drag as subversion, assumes a kind of singularity to drag queen identity, one that becomes interesting only in its potential to denaturalize heterosexual normativity. Queer linguistics, in contrast, invites us to discuss the conversational practices of all sexual identities - whether marginal or central to organizations of heterosexual kinship - as potentially felicitous on a more localized level. While much of the early research in the field has focused on the language practices of understudied sexual identities (just as much of the early research in language and gender focused on the language practices of women), its boundaries also embrace the findings of such scholars as Penelope Eckert (2002; Chapter 27 in this volume), whose ethnographic work on "the heterosexual marketplace" illustrates how heterosexual identity structures the adolescent social order in an American elementary school. Like queer theory, queer linguistics is necessarily concerned with how heterosexual normativity is produced, perpetuated, and resisted, but it seeks to localize these productions within specific communities of practice. (See Milani, Chapter 13 in this volume, for a discussion of this dimension of queer theory.) In the last decade, the field of queer linguistics has come under fire from Don Kulick (2000; Chapter 3 in this volume), who argues that the language practices of gays and lesbians must be "unique to gays and lesbians" (2000, 259) if they are to be of interest to sociolinguists. Kulick takes difference to be the necessary starting point for scholarship on language and sexuality, arguing that because linguistic dif­ ferences across sexual identities have not been satisfactorily demonstrated, the field is not viable. Yet an insistence on difference not only requires linguistic deviance as a prerequisite for sociolinguistic research; it also recalls the much criticized dif­ ference model of language and gender (see Bucholtz and Hall 2004 for a fuller discussion). This approach, as noted earlier, has been extensively problematized for its tendency to emphasize cross-gender variation at the expense of potentially more significant intragender variation and cross-gender similarity. The practice-based and ideological models of language and gender that developed in response to these critiques, such as queer linguistics, seek not to describe how women's lan­ guage use differs from men's, or how homosexuals' language use differs from heterosexuals', but to document the diverse range of women's and men's linguis­ tic repertoires as developed within particular contexts. In these models, gender is seen as materializing only in interaction with other sociological discourses, includ­ ing historical, national, ethnic, racial, age-related, and sexual ones. This, I would argue, is the direction that research on language and sexual identity must continue to take if the exceptional speakers of previous generations are to move squarely out of the footnotes.

Acknowledgments

This chapter is an abbreviated version of a longer article that appeared in the 2003 edition of theHandbook. I would like to express my thanks to editors Miriam Contested and Problematized Identities 237

Meyerhoff, Janet Holmes, and Susan Ehrlich for several careful readings of this chapter, and to Joshua Raclaw, Lai Zimman, and Chad Nilep, who provided excellent suggestions on ways to revise this chapter for the second edition.

REFERENCES

Austin, John L. 1962.How to Do Things with Burris, Barbara. 1973. "Fourth World Words. Cambridge, MA: Harvard Manifesto." In Anne Koedt, Ellen Levine, University Press. and Anita Rapone (eds.), Radical Feminism, Barrett, Rusty. 1999. "Indexing Polyphonous 322-357. New York: Quadrangle Books. Identity in the Speech of African American Butler, Judith. 1990.Gender Trouble: Feminism Drag Queens." In , Anita C. and the Subversion of Identity. New York: Liang, and Laurel A. Sutton (eds.), Routledge. Reinventing identities: The Gendered Self in Chamberlain, Alexander F. 1912. "Women's Discourse, 313-331. New York: Oxford Languages."American Anthropologist, 14: University Press. 579-581. Bauman, Richard, and Charles L. Briggs. Crosby, Faye, and Linda Nyquist. 1977. "The 1990. "Poetics and Performance as Critical Female Register: An Empirical Study of Perspectives on Language and Social Life." Lakoff's Hypotheses."Language in Society, Annual Review of Anthropology, 19: 59-88. 6: 313-322^ Blumstein, Philip, and Pepper Schwartz. Dubois, Betty Lou, and Isabel Crouch. 1975. 1984. American Couples: Money, Work, Sex. "The Question of Tag Questions in New York: William Morrow. Women's Speech: They Don't Really Use Bogoras, Waldemar. 1922. "Chukchee." In More of Them, Do They?"Language in Franz Boas (ed.), Handbook of American Society, 4: 289-294. Indian Languages 2. Bureau of American Dunbar, Roxanne. 1970. "Female Liberation Ethnology Bulletin 40, 631-903. as the Basis for Social Revolution." In Washington, DC: Smithsonian Institution. Robin Morgan (ed.),Sisterhood is Powerful, Bucholtz, Mary, and Kira Hall. 1995. "Twenty 477-492. New York: Random House. Years after Language and Woman's Place." In Kira Hall and Mary Bucholtz (eds.),Gender Eckert, Penelope. 2002. "Demystifying Articulated: Language and the Socially Sexuality and Desire." In Kathryn Constructed Self, 1-22. New York: Campbell-Kibler, Robert J. Podesva, Sarah Routledge. J. Roberts, and Andrew Wong (eds.), Bucholtz, Mary, and Kira Hall. 2004. Language and Sexuality: Contesting Meaning "Theorizing Identity in Language and in Theory and Practice, 99-110. Stanford: Sexuality Research."Language in Society, CSLI Publications. 33(4.): 501-547. Eckert, Penelope, and Sally McConnell-Ginet. Bucholtz, Mary, Anita C. Liang, and Laurel 1992. "Think Practically and Look Locally: A. Sutton, eds. 1999.Reinventing Identities: Language and Gender as The Gendered Self in Discourse. New York: Community-Based Practice."Annual Oxford University Press. Review of Anthropology, 21: 461-490. Burrell, Nancy, and Mary Anne Fitzpatrick. Flannery, Regina. 1946. "Men's and Women's 1989. "The Psychological Reality of Marital Speech in Gros Ventre."International Conflict." In Dudley D. Cahn (ed.). Journal of American Linguistics, 12: 133-135. Intimates in Conflict: A Communication Furfey, Paul Hanly. 1944. "Men's and Perspective, 167-186. Hillsdale, NJ: Women's Language."American Catholic Lawrence Erlbaum. Sociological Review, 5: 218-223. 238 Identities

Gal, Susan. 1991. "Between Speech and Lakoff, Robin. 1975.Language and Woman's Silence: The Problematics of Research on Place. New York: Harper & Row. Language and Gender/7 In Micaela di Lasch, Richard. 1907. "Über Sondersprachen Leonardo (eds.),Gender at the Crossroads of und ihre Entstehung."Mitteilungen der Knowledge: Feminist Anthropology in the Anthropologischen Gesellschaft in, 37: Wien Modern Era, 175-203. Berkeley: University 89-101. of California Press. Livia, Anna, and Kira Hall. 1997, " Tt's a Gatschet, Albert S. 1884.Hitchiti: A Migration Girl!7 Bringing Performativity Back to Legend of the Creek Indians. Philadelphia: Linguistics." In Anna Livia and Kira Hall Brinton. (eds.), Queerly Phrased: Language, Gender, Greenough, James Bradstreet, and George and Sexuality, 1-18. New York: Oxford Lyman Kittredge. 1901.Words and Their University Press. Ways in English Speech. New York: Maccoby, Eleanor E/1998.The Two Sexes: Macmillan. Growing Up Apart, Coming Together. Gumperz, John J., ed, 1982.Language and Cambridge, MA: Harvard University Social Identity. Cambridge: Cambridge Press. University Press. Maltz, Daniel N., and Ruth A. Borker. 1982. Haas, Mary R. 1964 [1944]. "Men's and "A Cultural Approach to Male-Female Women's Speech in Koasati.77 Language, 20: Miscommunication.77 In John J. Gumperz 142-149. Repr. in Dell Hymes (ed.), (ed.). Language and Social Identity, 196-216. Language in Culture and Society, 228-233. Cambridge: Cambridge University Press. New York: Harper & Row. McElhinny, Bonnie. 2002. "Language, Hall, Kira. 1999. Performativity.Journal of Sexuality, and Political Economy." In , 9(1-2): 184-187. Kathryn Campbell-Kibler, Robert J. Hymes, Dell. 1975. "Breakthrough into Podesva, Sarah J. Roberts, and Andrew Performance." In Dan Ben-Amos and Wong (eds,), Language and Sexuality: Kenneth S. Goldstein (eds.),Folklore: Contesting Meaning in Theory and Practice, Performance and Communication, 11-74. 111-134. Stanford: CSLI Publications. The Hague: Mouton. Nicholson, Linda. 1994. "Interpreting Jacobs, Greg. 1996. "Lesbian and Gay Male Gender." Signs, 20(1): 79-105. Language Use: A Critical Review of the O'Barr, William M., and Bowman K. Atkins. Literature."American Speech, 71(1): 1980. " 'Women's Language7 or 'Powerless 49-71. Language7?" In Sally McConnell-Ginet, Jespersen, Otto. 1990 [1922]. "The Woman." Ruth Borker, and Nelly Furman (eds.), In Language: Its Nature, Development, and Women and Language in Literature and Origin. London: Allen & Unwin. Repr. in Society, 93-109. New York: Praeger. Deborah Cameron (ed.),The Feminist Sapir, Edward. 1949 [1915]. "Abnormal Types Critique of Language: A Reader, 201-220. of Speech in Nootka." Reprinted in David New York: Routledge. Mandelbaum (ed.), Selected Writings of Keenan (Ochs), Elinor. 1996 [1974]. Edward Sapir, 179-196. Berkeley: "Norm-Makers, Norm-Breakers: Uses of University of California Press. Speech by Men and Women in a Malagasy Sapir, Edward. 1949 [1929]. "Male and Community." Repr. in Donald Brenneis Female Forms of Speech in Yana." Repr. in and Ronald K. S. Macaulay (eds.),The David Mandelbaum (ed.), Selected Writings Matrix of Language: Contemporary Linguistic of Edward Sapir, 206-212. Berkeley: Anthropology, 99-115. Boulder, CO: University of California Press. Westview Press. Tánnen, Deborah. 1986.That's Not What Kulick, Don. 2000. "Gay and Lesbian I Meant! How Conversational Style Makes Language. "Annual Review of Anthropology, or Breaks Relationships. New York: 29: 243-285. Ballantine. Contested and Problematized Identities 239

Tannen, Deborah. 1990.You Just Don't Liang, and Laurel A. Sutton (eds.), Understand: Women and Men in Reinventing Identifies: The Gendered Self in Conversation. New York: Ballantine Books. Discourse, 101-122. New York: Oxford Thorne,Barrie. 1993. Gender Play: Girls and University Press. Boys in School. New Brunswick, NJ: Wolf,Arthur. 1995. Sexual Attraction and Press. Childhood Association: A Chinese Brief for Trechter, Sara. 1999. "Contextualizing the Edward Westermarck. Stanford: Stanford Exotic Few." in Mary Bucholtz, Anita C. University Press.